How much do Facebook ads cost in 2026?

By Sofia Marchetti, co-founder & measurement lead · Published November 6, 2025 · Updated August 19, 2026

The honest answer has ranges, not a number. In the freshest panels: ecommerce accounts paid a median $15.06 CPM over the year to July 2026, US advertisers paid $0.70 per click on traffic campaigns and $1.92 on leads campaigns, and the median ecommerce cost per acquisition was $38.99. Any single "average Facebook ads cost" you have read is a real number from one panel, quietly presented as the whole market.

Below is what each cost metric runs right now, who measured it, and what moves it. The full tables live on our benchmarks page, which we re-verify quarterly.

What does 1,000 impressions cost (CPM)?

Three defensible answers exist at once. Triple Whale's ecommerce panel of 40,000+ stores shows a $15.06 median for the year to July 2026, up 13% year over year. The Ryze cross-vertical panel averages $14.19 for 2026. Gupta Media's own buying averaged $8.19 across 2025, dipping near $6.59 in October before the holiday surge. The spread is not measurement error: different verticals, objectives, and optimization events genuinely cost different amounts to serve.

Vertical is the biggest lever. Education advertisers see CPMs near $7.60 while insurance pays around $21.40, nearly a threefold difference before anyone touches a campaign setting.

What does a click cost (CPC)?

Objective decides. WordStream's US data, covering 554 traffic campaigns and 726 leads campaigns, shows $0.70 per traffic click against $1.92 per leads click. That gap is rational: leads campaigns bid toward people who fill in forms, and the auction prices that intent. By industry, clicks range from $0.45 in apparel to $3.77 in finance.

Treat CPC as a diagnostic, never a target. A cheap click that never converts is expensive; we have watched accounts chase $0.50 clicks straight into a $90 CPA.

What does a customer cost (CPA and CPL)?

The ecommerce median CPA sits at $38.99 for the year to July 2026. For lead generation, the cross-industry average cost per lead reached $27.66 in the 2025 WordStream panel, up 21% in a single year. Industry ranges run from $7.85 to $198.42, which is why a franchise comparing itself to a newsletter signup funnel learns nothing. Our own accounts median a 27% CPA reduction in the 90 days after restructure; the counting rules behind that number are published on the track record page.

When do costs spike?

Q4, reliably and dramatically. The Gupta Media tracker had weekly Facebook CPMs peaking at $13.42 over Thanksgiving and Black Friday 2025 against the $8.19 annual average, a 64% premium at minimum. Auction pressure starts building in early November and releases just after mid-December. Plans that hold budgets flat through that window buy fewer impressions each week and then blame the creative.

What actually moves your costs?

Four things, in rough order of force. Creative quality: Meta's own analysis attributes about 56% of auction outcomes to it. Event data quality: delivery optimizes on what you send, and a Pixel plus Conversions API setup with strong match quality is worth roughly 13% lower cost per action by Meta's published figures. Account structure: consolidated setups clear the learning phase faster and waste less. Seasonality, as above. Audience hacking is conspicuously absent from that list; since Andromeda, targeting is mostly the machine's job.

Why do prices keep rising when there are more ads than ever?

Because demand grows faster than inventory, and the inventory keeps getting better at converting. Meta booked $196.18 billion in advertising revenue in 2025, up 22% in a year, from a pool it last officially sized at more than 10 million active advertisers. In the second quarter of 2026, ad impressions rose 14% while the average price per ad still climbed 12%: more slots and higher prices at the same time, which only happens when advertisers keep finding the slots worth it.

eMarketer projects Meta will pass Google as the world's largest ad seller in 2026. For your budget, the practical translation is that CPM relief is not coming. Efficiency has to come from the inputs you control: creative, event quality, and structure.

How do you know if YOUR costs are the problem?

Run three comparisons before blaming the market. Your CPM against your industry band (not the global median). Your CTR against the 1.71 to 2.59% objective range, because a CTR far below band usually means creative, not auction pressure. And your CPA trend against your CPM trend: if CPM rose 13% and your CPA rose 40%, the auction explains a third of your problem and the account explains the rest. That decomposition is the first page of the audit, and it settles most cost arguments in one table.

How should you budget from these numbers?

Work backward from unit economics. Take your contribution margin per sale, decide the CPA you can survive, and check it against your industry band rather than the global median. Then fund at least 50 conversions a week per ad set so delivery can learn. If the arithmetic breaks below $3,000 a month of spend, run the numbers again before running the ads; the answer might be margin work, not media.